Skip to content

jhasan2026/LangGraphResearchLab

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🔥 Mastery Path: 500 LangChain + LangGraph Practice Problems


🧱 SECTION 1 — Foundations & Hardcoded Workflows (1–40)

Sequential (Hardcoded)

  1. Sequential node pipeline that adds two numbers
  2. BMI calculator (hardcoded inputs)
  3. Temperature conversion workflow (C → F → advisory)
  4. Loan eligibility calculator
  5. Tax slab calculator
  6. Grade calculator from marks
  7. Salary breakdown pipeline
  8. Currency conversion (static rates)
  9. Interest calculator (simple → compound)
  10. Age-based recommendation system

Parallel (Hardcoded)

  1. Parallel node for sum, avg, max
  2. Cricket score calculator (runs + strike rate)
  3. Election result aggregation
  4. Weather stat aggregation
  5. E-commerce cart price breakdown
  6. Parallel validation of form fields
  7. Parallel risk scoring
  8. Parallel fraud flags
  9. Parallel sensor readings
  10. Parallel exam result processing

Conditional (Hardcoded)

  1. Quadratic equation solver
  2. If/else grade assignment
  3. Loan approval decision tree
  4. Traffic signal logic
  5. Discount eligibility logic
  6. Health risk classifier
  7. Shipping cost decision graph
  8. Student pass/fail logic
  9. Admission eligibility
  10. Insurance premium decision

Iterative (Hardcoded)

  1. Loop until number becomes prime
  2. Retry until valid input
  3. Iterative average convergence
  4. Looping counter graph
  5. Fibonacci generator
  6. Iterative sorting simulation
  7. Guess-the-number loop
  8. Retry-on-failure workflow
  9. Countdown timer graph
  10. Repeated validation loop

🧠 SECTION 2 — First LLM Integration (41–100)

Sequential + LLM

  1. Question → Answer pipeline
  2. Prompt → summary → title
  3. Resume → improvement → final output
  4. Blog outline → blog generation
  5. Email draft → tone refinement
  6. Question rewriter → answerer
  7. User intent → response
  8. Translation pipeline
  9. Text cleanup → summarization
  10. News → sentiment → summary

Parallel + LLM

  1. Essay grammar + coherence + score
  2. Resume skill extraction + scoring
  3. Review sentiment + category
  4. Multi-perspective answer generator
  5. Parallel language translations
  6. Parallel keyword extraction
  7. Parallel fact checking (mock)
  8. Parallel explanation styles
  9. Parallel answer evaluation
  10. Multi-metric content scoring

Conditional + LLM

  1. Review reply generator
  2. Toxicity detection → safe reply
  3. Question difficulty classifier
  4. Intent-based routing chatbot
  5. Email urgency classifier
  6. Support ticket router
  7. Resume rejection/acceptance
  8. Fake news detector (LLM-based)
  9. Query complexity classifier
  10. Customer churn risk responder

Iterative + LLM

  1. Iterative tweet improvement
  2. Retry until score > threshold
  3. Loop until hallucination reduced
  4. Self-refinement loop
  5. Debate loop (pro vs con)
  6. Multi-round explanation refinement
  7. Iterative summarization compression
  8. Answer critique → regenerate
  9. Prompt optimization loop
  10. Language simplification loop

🤖 SECTION 3 — Chatbots & Memory (101–160)

Basic Chatbots

  1. Stateless chatbot
  2. Chatbot with buffer memory
  3. Chatbot with summary memory
  4. FAQ chatbot
  5. Rule-based chatbot
  6. Customer support bot
  7. Interview chatbot
  8. Tutor chatbot
  9. Product recommendation bot
  10. Mental health support bot (safe)

Graph-Based Chatbots

  1. Intent-routing chatbot
  2. Multi-agent chatbot
  3. Tool-using chatbot
  4. Memory-aware chatbot
  5. Context-aware chatbot
  6. Conditional response chatbot
  7. Escalation-based chatbot
  8. Multi-language chatbot
  9. Conversation-state graph
  10. Topic-switching chatbot

Iterative Chat

  1. Clarifying-question bot
  2. Answer verification bot
  3. Follow-up generator
  4. Debate chatbot
  5. Negotiation chatbot
  6. Interview feedback bot
  7. Teaching chatbot
  8. Code-review chatbot
  9. Storytelling chatbot
  10. Therapy-style reflective bot

📚 SECTION 4 — Retrieval & RAG Graphs (161–230)

Basic RAG

  1. PDF QA bot
  2. Website QA bot
  3. Notes QA system
  4. Company policy QA
  5. Legal document QA
  6. Research paper QA
  7. Medical FAQ bot
  8. Book QA system
  9. Course material QA
  10. Knowledge base QA

RAG with Graph Logic

  1. Query → retrieve → answer → verify
  2. Confidence-based retrieval
  3. Conditional fallback to LLM
  4. Multi-source retrieval
  5. Citation-enforced RAG
  6. RAG + summarization
  7. RAG + explanation
  8. RAG + comparison
  9. RAG + contradiction check
  10. RAG + hallucination guard

Iterative RAG

  1. Query reformulation loop
  2. Retrieve until confidence met
  3. Chunk-size experimentation
  4. Multi-hop retrieval
  5. Evidence refinement loop
  6. Self-critique RAG
  7. Answer verification loop
  8. Retriever ranking loop
  9. Context compression loop
  10. Dynamic retrieval depth

🧩 SECTION 5 — Agents & Tools (231–300)

Tool Use

  1. Calculator tool agent
  2. Web search tool agent
  3. File reader agent
  4. Email drafting agent
  5. Code execution agent
  6. Math solver agent
  7. Data analysis agent
  8. SQL query agent
  9. API caller agent
  10. Markdown generator agent

Multi-Agent Graphs

  1. Researcher + Writer
  2. Planner + Executor
  3. Critic + Generator
  4. Teacher + Student
  5. Analyst + Summarizer
  6. Debate agents
  7. Reviewer + Fixer
  8. Translator + Verifier
  9. Security checker + responder
  10. Manager + workers

Agent Loops

  1. Agent retries until success
  2. Tool selection loop
  3. Self-debugging agent
  4. Multi-step planning agent
  5. Autonomous task agent
  6. Agent with memory
  7. Agent with reflection
  8. Agent with scoring
  9. Agent with human-in-the-loop
  10. Agent with budget constraints

🏗️ SECTION 6 — Complex Graph Architectures (301–380)

Advanced Graph Patterns

  1. DAG workflow
  2. Cyclic graph
  3. Dynamic node creation
  4. State mutation graph
  5. Subgraph composition
  6. Hierarchical graphs
  7. Event-driven graph
  8. Streaming graph
  9. Timeout-handling graph
  10. Error-recovery graph

Business Use Cases

  1. Hiring pipeline
  2. Loan processing system
  3. E-commerce assistant
  4. Marketing campaign generator
  5. HR policy chatbot
  6. SaaS onboarding assistant
  7. Financial advisor bot
  8. Legal assistant
  9. Healthcare triage bot
  10. Education advisor

Evaluation & Monitoring

  1. Output scoring node
  2. Hallucination detector
  3. Toxicity filter
  4. Confidence scorer
  5. Latency monitor
  6. Cost tracking graph
  7. Prompt drift detection
  8. Failure logging
  9. Retry strategy graph
  10. Human feedback integration

🚀 SECTION 7 — Research-Level & Mastery Problems (381–500)

Research-Oriented

  1. Graph-based prompt optimization
  2. Auto-discover optimal graph
  3. Self-improving workflow
  4. Meta-agent controller
  5. Curriculum learning graph
  6. Multi-objective optimization
  7. Graph ablation study
  8. Graph explainability
  9. Graph performance benchmarking
  10. Graph visualization tool

Safety & Alignment

  1. Safety-aware chatbot
  2. Bias detection graph
  3. Jailbreak prevention workflow
  4. Ethical response filter
  5. PII detection system
  6. Red-team testing graph
  7. Compliance checker
  8. Policy enforcement agent
  9. Risk-aware generation
  10. Trust score computation

Capstone Systems

  1. Autonomous research assistant
  2. AI teaching assistant
  3. Startup co-founder agent
  4. Scientific paper reviewer
  5. Grant proposal generator
  6. Multi-modal RAG system
  7. Codebase understanding agent
  8. AI OS task manager
  9. Knowledge graph builder
  10. Long-running AI service

Extreme Mastery

  1. Self-debugging LangGraph
  2. Auto-generated LangGraph
  3. Prompt-to-graph compiler
  4. Natural language → workflow
  5. Graph mutation at runtime
  6. Distributed LangGraph
  7. Multi-user graph memory
  8. Graph checkpointing system
  9. Graph rollback mechanism
  10. Graph A/B testing

Final Boss Problems

421–500. Rebuild everything above but with:

  • Streaming
  • Human-in-the-loop
  • Cost constraints
  • Safety guarantees
  • Multi-agent coordination
  • Evaluation baked in
  • Production readiness

About

A complete 500-problem roadmap to master LangChain and LangGraph — from basic workflows to advanced agents, RAG systems, multi-agent graphs, safety, evaluation, and production-ready AI orchestration.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors